I think that to make a statement like "conceptualizing the nature of the universe's 'edge' is irrelevant because it's unobservable" and refusing to even consider the possibilities is rather unproductive, if not downright unscientific. Through simulation and modeling we can construct hypotheticals, and quite accidentally run across something useful or testable. Just because we can't fly a probe to the edge of the cosmos doesn't mean we won't construct a model that just so happens to explain, say, lambda/the cosmological constant/whatever as an added bonus. Many times, a theory been made to explain one thing - and then, quite accidentally, someone steps back and says 'hey! this also explains THIS wacky phenomenon!' They may never experimentally verify the thing they originally set out to explain, but those happy accidentals that end up getting explained as a result certainly can go a ways in supporting the theory.
I get sensitive about those sorts of statements because they half-seem to be actively trying to stifle progress. String theory, for instance, predicts nothing, and instead of simply being content to simply not believe it, some people go out of there way to call it useless, a waste of time, etc. Those people will be in for a surprise if ever some testable prediction does pop up at some point as a result of the work that's been put into it (Note, I'm not trying to turn this into a string theory discussion, I just used it as an example).
In any case, it's a fun question whether verifiable or not. Now, to get more on topic:
When I was 15, I was puzzing over this same question (or one similar enough). Assuming the universe is a finite bubble of spacetime, what happens to the traveller that sneaks up to the 'edge' and tries to push beyond? Is there a beyond?
I followed a few assumptions (and note they're only assumptions):
The universe is
- finite
- spherical
- spacetime is the 'medium' in which matter, energy, and measureable dimensions exist
- anything 'outside' of the bubble of spacetime must have zero measurable distance. There is no 'there' there
Okay. The universe obviously has a measurable diameter, and a volume. Here's where I ran into a puzzler: How can a universe have an edge, a perimeter, and outside surface area as it were, and be 'surrounded' by an area of zero distance? I figured that was impossible - that a traveller moving toward the outside edge would be getting closer and closer to an area of zero distance. There can be no such thing as a 'perimeter' or surface area to the bubble if it's rubbing elbows, or expanding into, with an 'area' of zero distance.
But the universe does have pockets of measurable area that 'rub up' against areas of (theoretically) zero distance: singularities.
Therefore, as one would travel to the 'edge' of the universe, he'd find that he's moving through space that curves into a singularity. That's the only way the universe can be 'surrounded' by an area of zero distance - no matter what direction you travel in, go far enough and you'll fall into a black hole, basically. This means the universe has a diameter and an area, but no perimeter.
It would also mean that the 'edge' of the universe is a massive gravity well. :)
I invented a word for this all-encompassing singularity; 'pangularity', from the latin pan meaning 'all'. (It was easier to pronounce than 'omnigularity'.)
This tied into something else I was pontificating about at the same time: relativity in a spinning disc. If you take a disc and spin it so the outer edge accelerates near the speed of light, the measured perimeter would shrink, but the diameter and area remains the same. If it was a magical disc whose edge could spin at the speed of light, you'd have a perimeter of zero distance. Given the equivalence of acceleration and gravity, the relativistic spinning disc could serve (at least mathematically) as a two-dimensional model for the pangularity idea. This was good news for me, because I was failing Algebra at the time. :)
I ran into a problem. At the time, it was 'common knowledge' that the universe's expansion was slowing. According to my hypothesis, the further you traveled away from the center of the universe, you should start falling into the 4D curvature of space into the pangularity, and therefore start speeding up instead of slowing down. So I said 'well, so much for that', congratulated myself for having a neat idea and forgot all about it, until I read an article about a year later about the surprising new measurements that demonstrated the universe was accelerating in its expansion. I did a happy jig and congratulated myself for being the smartest human being on the planet.
I started getting silly and conjecturing completely baseless speculations, such as the existence of a white hole, a convex singularity, at the center of the universe serving as the 'other side' of the pangularity, and the universe constantly recycling its matter and energy... etc.
Of course, that was almost 15 years ago, and I've gotten a little wiser, if not smarter. I don't think it can be adapted to explain why everything is moving apart from everything else at an accelerating rate. As two objects get closer to the outer edge of the universe, they should actually come closer together as they fall into the pangularity. And if the gravity well's effects were to extend into our local observable area of the universe, movement of objects should seem biased in one direction, assuming we're not parked in the exact center of the universe, etc. Of course, the volume and diameter of the universe could still be expanding due to inflation in this model, but the perimeter is always zero.
I still suspect that if spacetime is expanding 'into nothing', then the perimeter must have zero distance.
But it was fun speculation, and speculation can lead to some very interesting ideas. I think that the culture of strict positivism ('the question is irrelevant!', etc) seems counterproductive.